{
  "id": 55534,
  "title": "Apple's Approach to Speaker-Independent Recognition",
  "url": "/competitions/freesound-audio-tagging/discussion/55534",
  "author_name": "Humza Ahmed",
  "post_date": "2018-04-28T04:03:47.450000",
  "votes": 3,
  "comment_count": 0,
  "views": 0,
  "content": "<p>I thought that the latest machine learning journal post by Apple about recognizing and personalizing Siri may be of interest to those competing in this challenge. </p>\n\n<p>Journal Post Here:</p>\n\n<p><a href=\"https://machinelearning.apple.com/2018/04/16/personalized-hey-siri.html\">https://machinelearning.apple.com/2018/04/16/personalized-hey-siri.html</a></p>\n\n<p>Looks like the team represents audio using a transform incorporating both MFCC + HMM to recognize \"Hey, Siri\". </p>\n\n<p>Feel free to discuss below!</p>",
  "messages": [
    {
      "id": 320263,
      "postDate": "2018-04-28T04:03:47.450Z",
      "content": "<p>I thought that the latest machine learning journal post by Apple about recognizing and personalizing Siri may be of interest to those competing in this challenge. </p>\n\n<p>Journal Post Here:</p>\n\n<p><a href=\"https://machinelearning.apple.com/2018/04/16/personalized-hey-siri.html\">https://machinelearning.apple.com/2018/04/16/personalized-hey-siri.html</a></p>\n\n<p>Looks like the team represents audio using a transform incorporating both MFCC + HMM to recognize \"Hey, Siri\". </p>\n\n<p>Feel free to discuss below!</p>",
      "rawMarkdown": "I thought that the latest machine learning journal post by Apple about recognizing and personalizing Siri may be of interest to those competing in this challenge. \n\nJournal Post Here:\n\nhttps://machinelearning.apple.com/2018/04/16/personalized-hey-siri.html\n\nLooks like the team represents audio using a transform incorporating both MFCC + HMM to recognize \"Hey, Siri\". \n\nFeel free to discuss below!",
      "votes": 3
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "320263": "I thought that the latest machine learning journal post by Apple about recognizing and personalizing Siri may be of interest to those competing in this challenge. \n\nJournal Post Here:\n\nhttps://machinelearning.apple.com/2018/04/16/personalized-hey-siri.html\n\nLooks like the team represents audio using a transform incorporating both MFCC + HMM to recognize \"Hey, Siri\". \n\nFeel free to discuss below!"
  }
}